Speaker
Description
Open educational materials for data analysis and programming often depend on datasets that need to remain accessible, well-documented, and reusable over time. At the iBehave open technology support (iBOTS), we develop such materials for neuroscience researchers. This raises several research data management challenges: How can datasets be made available through stable public references while allowing external contributions? How can their structure and content be validated and versioned? How can datasets be linked to the educational materials that depend on them? And how can such a system be scaled to larger catalogs of materials while keeping infrastructure costs and maintenance effort low?
We present a preliminary workflow that combines existing open infrastructure and standards. We use the Data Package standard for schema-based dataset validation, Zenodo communities for publication and distribution, cloud object storage for backups, and custom Python tooling for managing dataset versions and their relationships to educational materials. The aim is to support FAIR (findable, accessible, interoperable, reusable) principles while keeping the workflow lightweight and maintainable.
Our poster presents the architecture, the design decisions behind it, and the challenges that remain. We hope to exchange experiences with educators, research data managers, and Open Science practitioners on approaches to managing reusable datasets and on how similar workflows could support accessible and reproducible research and training materials across disciplines.